Gustavo Bakker

Born in the Netherlands but currently living in Portugal, I’m an experienced Software Developer who is passionate about Artificial Intelligence.

I am a graduate from ISCTE - Lisbon University Institute in Telecommunications and Computing Engineering and I have 4 years experience as an IT consultant building scalable Java web applications with Spring framework for big international companies.

My taste for AI begun some years ago in the academic environment when I started building an autonomous program to draw/paint different patterns and colors in a 2D matrix, with the goal of creating interesting artistic drawings. Then I reused that code to create a version of the well-known Snake game, using the A* shortest path algorithm to create an AI that allows the snake to autonomously eat the apples and avoid every possible obstacle (walls) in the way. Unfortunately it did not solve things like the Hamiltonian path problem in which the snake avoids to overlap itself. I implemented a program that uses OpenCV library to perform face detection. Also did a Java chatbot based on AIML with the goal of making a bot that understands the context of a conversation.

Few years later I started to read about Machine Learning and I realized the impact this field was producing in AI, so I redirected my focus from the old fashioned AI algorithms to this field. The Machine Learning course from Coursera was the best way to get into it and understand the basics. Now I try to dive deeper into more practical stuff using Python, practicing and building real-world predictive models with Scikit-learn, Keras, NumPy and Pandas libraries. Currently I like to build and optimize different kinds of supervised learning models such as SVM, Decision Trees, Random Forests and Neural Networks (CNN and RNN), using frameworks such as Tensorflow and PyTorch, and also unsupervised learning models, such as K-means, Self-Organized Maps, Boltzmann Machines and Autoencoders.

Recently, I started to work with Reinforcement Learning, namely Deep Q-learning, as my new tool to solve AI challenges in games and creative programs.

Currently, my effort and motivation goes mainly to Deep Learning and Reinforcement Learning applications.

This is a place where I share some of my ideas.

Stay tunned.